Polarforschung 72 (2/3), 79 - 92, 2002 (erschienen 2004)

Marine Mammal Auditory Systems:A Summary ofAudiometric and Anatomical Dataand Implications for Underwater Acoustic Impacts

by Darlene R. Ketten'

TERMINOLOGY

Audiogram: A graph of hearing ability conventionally dis­played as frequency (abscissa) versus sensitivity measuredas sound pressure 01' intensity (ordinate).

Cetaceans: Wh ales and dolphins.decibel (dB): A scale based on the log ratio of two quantities.

lt is commonly used to represent sound pressure level. Thevalue of the decibel depends upon the reference pressureused. Therefore the decibel level of sound is properly statedin the form of n dB re n microPa. The ,uPa is a unit of pres­sure. In terms of intensity, 100 dB re 20 ,LIPa in air equalsapproximately 160 dB re 1 fiPa in water.

infrasonic: Below 20 Hz, the lower limit of human hearing.kHz: kilo Hertz. A Hertz (Hz) is a measure of sound fre­

quency equal to I cycle sec'. A kHz is one thousand cyclesper second.

Mysticetes: Baleen or moustached whales, which include thelargest whales such as Blue and Finback Whales are notknown to echolocate.

Octave: An octave is broadly defined as a doubling of fre­quency. Thus, a one octave shift from 500 Hz is 1,000 Hz;from 3,000 Hz, it is 6,000 Hz. Adult humans have onaverage an eight octave functional hearing range of 32 Hz to16 kHz.

Odontocetes: Toothed whales; all are believed to echolocate;i.e., to use asound for imaging the environment.

INTRODUCTION

Concomitant with man's increasing use of the oceans is anincrease in the ocean's acoustic budget. In the mid 1970 's, itwas estimated that noise from human related activity wasincreasing in coastal areas and shipping lanes at 10 dB perdecade. Given our ever increasing activity in all seas and at alldepths, this figure is not surprising. It may even be too conser­vative. Anthropogenie noise is an important component ofvirtually every human endeavour in the oceans, whether it beshipping, transport, exploration, research, military activities,construction, 01' recreation. For some activities, such as mili­tary and construction, impulsive and explosive devices arefundamental tools that are intermittent but intense; for others,such as shipping, the instantaneous noise may be less, butsound is inherent in daily operations and is therefore a

, Woods Hole Occanoraphic Institution, Departmcnt of Biology, Woods Hole, MA02543, USA. and Department of Otology and Laryngology, Harvard Medical School;<dketten@whoi.cdu>

constant, pervasive by-product. Because these activities spanthe globe and produce sounds over the entire audible range ofmost animals, it is reasonable to assume that man-made noisein the oceans can have a significant adverse impact on marinespecies. Because marine mammals are especially dependentupon hearing and in many cases are endangered, the concernover noise impacts on these animals is particularly acute. Ourconcern is both logical and appropriate, but it is also, at thistime, unproved and the range of concerns is unbounded. Forresponsible stewardship of our oceans it is imperative that webegin to measure and understand our impacts, and, moreimportant, that we proceed with a balanced and informedview. To that end, this meeting is a significant, positive step.

Hearing for any animal is an important sense. Many sensorycues are limited in their distribution and utility. Sound how­ever is literally universal. While many animals inhabit light­less environments and are blind, there are no knownvertebrates that are naturally, profoundly deaf. There is nohabitat, except space, that is soundless, and sound is such asignificant cue, carrying such a wealth of information thathearing is weil developed in virtually every animaI group. Weemploy sound and hearing both passively and actively,listening not only in the dark but even while asleep. The cuesare constant and diverse, providing information on the direc­tion and nature of the sources and how they change throughtime. Sound is a key element for survival and hearing is a keycomponent of communication, mate selection, feeding, andpredator avoidance

For marine mammals, hearing is arguably their premier sen­sory system. lt is obvious from their level of ear and neuralauditory center development alone. Dolphins and whalesdevote three-fold more neurons to hearing than any otheranimal. The temporal lobes, which control higher auditoryprocessing, dominate their brain, and they appear to havefast er auditory and signal processing capabilities than anyother mammal. Since the late 1950's we have been aware thatdolphins, at least, use very high ultrasonic signals as a form ofbiosonar. Using sound they can distinguish amongst differentmetals and detect differences as small as a few mm in twoobjects. To date, despite 50 years of research on dolphinbiosonar, we are still incapable of duplicating some of theirfeats. However, despite the multifaceted evidence we have forexceptional and diverse hearing in marine mammals, we stillknow very little about how and what they hear.

This statement summarizes and critiques existing auditorydata for marine mammals, It was compiled primarily as a

79

background document for assessing potential impacts ofanthropogenic sounds, including long-range detection 01' sonardevices. To that end, it has the following emphases: a descrip­tion of currently available data on marine mammal hearing andear anatomy, a discussion and critique of the methods used toobtain these data, a summary and critique of data based onhearing models for untested marine species, and a discussionof data available on acoustic parameters that induce auditorytrauma in both marine and land mammals. In order to placethese data in an appropriate context, summaries are incorpo­rated also of basic concepts involved in underwater versus air­borne sound propagation, fundamental hearing mechanisms,and mechanisms of auditory trauma in land mammals. Lastly,to maximize the utility ofthis document, abrief discussion hasbeen included on the potential for impact on hearing fromseveral recently proposed devices and an outline of researchareas that need to be addressed if we are to fill the relativelylarge gaps in the existing data base

MAMMALIAN HEARING FUNDAMENTALS

The term "auditory system" refers generally to the suite ofcomponents an animal uses to detect and analyze sound. Thereare two fundamental issues to bear in mind for the auditory asweIl as any sensory system. One is that sensory systems andtherefore perception are species-specific, The ear and what itcan hear is different for each species. The second is that theyare habitat dependent. In terms of hearing, both of these areimportant issues

Concerning the first issue, species sensitrvities, all sensorysystems are designed to allow animals to receive and processinformation from their surroundings which means they act ashighly selective filters. If every environmental cue availablereceived equal attention, the brain would be barraged by sen­sory inputs. Instead, sensory organs are essentially multilevelfilters, selecting and attending to signals that, evolutionarily,proved to be important.

Most animals have vocalizations that are tightly linked to theirpeak hearing sensitivities in order to maximize intraspecificcommunication, but they also have hearing beyond that peakrange that is related to the detection of acoustic cues frompredators, prey, 01' other significant environmental cues.Consider, in general, how predator and prey are driven to beboth similar and different sensorially. Because their activitiesintersect in place and time, they need, for example, to havesimilar visual and auditory sensitivities, but, ideally, differentfields ofview and hearing ranges. Similarly, two species livingwithin similar habitats 01' having common predators and preyhave some hearing bands in common but will differ in totalrange because of anatomical and functional differences thatare species dependent and reflect other "species-specific"needs. Thus, each animal 's perceived world is a differentsubset of the real physical world; it is a species-specificmodel, constructed from the blocks of data its particularsensory system can capture and process. Two species may haveoverlapping hearing ranges, but no two have identical sensitiv­ities. This is of course the case with piscivorous marinemammals, their fish targets, and with their prey competitors. Itis also the case with whales and ships. They both have naviga­tional and predator detection needs.

80

In animal behaviour, this concept is called the "Umwelt" (VONUEXKÜLL 1934). As a technica1 term, "Umwelt" means ananima1's perceptually limited construct of the world. Incommon usage, it means simply the environment. This dualmeaning reflects the complex interaction of sensory adapta­tions and habitat, which leads us to the second issue, i.e., therelation 01' influence of habitat on sensory abilities. Whilesenses are tuned to relevant stimuli by evolution they are nev­ertheless limited by the physical parameters ofthe habitat.

Mechanistically, hearing is a relatively simple chain of events:sound energy is converted by bio-mechanical transducers(middle and inner ear) into electrical signals (neural impulses)that provide a central processor (brain) with acoustic data.Mammalian ears are elegant structures, packing over 75,000mechanical and electrochemical components into an averagevolume of 1 crn', Variations in the structure and number of earcomponents account for most of the hearing capacity differ­ences among mammals.

Hearing ranges and the sensitivity at each audible frequency(threshold, 01' minimum intensity required to hear a givenfrequency) vary widely by species. "Functional" hearing refersto the range of frequencies a species hears without entrainingnon-acoustic mechanisms. In land mammals, the functionalrange is generally considered to be those frequencies that canbe heard at thresholds of 60 dB SPL, a decibel measure ofsound pressure level. The basis for this measure and how itdiffers in air and water is explained in the next section

By example, a healthy human ear has a potential maximumfrequency range of 0.02 to 20 kHz but the normal functionalhearing range in an adult is closer to 0.040 to 16 kHz (Fig. 1).In humans, best sensitivity (lowest thresholds) occurs between500 Hz and 4 kHz, which is also where most acoustic energyof speech occurs (SCHUKNECHT 1993, YOST 1994). Soundsthat are within the functional range but at high intensities(beyond 120 dB SPL) will generally produce discomfort andeventually pain. To hear frequencies at the extreme ends of anyanimal 's total range generally requires intensities that areuncomfortable, and frequencies outside 01' beyond our hearingrange are simply undetectable because of limitations in theear's middle and inner ear transduction and resonance charac­teristics. Through bone conduction 01' direct motion of theinner ear, exceptionally loud sounds that are outside the func­tional range ofthe normal ear can sometimes be perceived, butthis is not truly an auditory sensation.

"Sonic" is an arbitrary term derived from the maximal humanhearing range. Frequencies outside this range are deemedinfrasonic (below 20 Hz) 01' ultrasonic (above 20 kHz) sonic.We know that many animals hear sounds inaudible to humans;consider the training whistles in common use that are silent tohumans but clearly audible by dogs. Most mammals havesome ultrasonic hearing (i.e., can hear weIl at frequencies >20kHz) and a few, like the Asian Elephant, Elephas maximus,hear and communicate with infra-sonie signals «20 Hz)

That brings us to three major auditory questions:(1) What are the differences between marine and land mammalears?(2) How do these differences relate to underwater hearing?(3) How do these differences affect the acoustic impacts?

Fig. 1: Zones of hearing versus potential impactareas are shown for human hearing. The bottomcurve shows the average human threshold in airvs. frequency (YOST 1994). The white zone re­presents the generaJly safe zone. The grey zonerepresents the region in which temporary hear­ing loss is likely, but depends upon a combina­tion of intensity vs. length of exposure. Note thatthe border for probable onset of temporarythreshold shift is generaJly 80 dB over minimumthreshold and essentiaJly paraJleIs the normalhuman hearing curve. Discomfort and pain areby contrast essentiaJly flat functions independentof hearing threshold, with onsets near 120 dB re20 flPa (approximately 182 dB re I flPa equiva­lent).

To address these questions requires assimilating a wide vari­ety of data. Behavioural and electrophysiological measures areavailable for some odontocetes and pinnipeds, but there are nopublished hearing curves for any mysticete. We have anatom­ical data on the auditory system for approximately one-thirdof all marine mammal species, including nearly half of thelarger, non-captive species. These data allow us to estimatehearing based on physical models of the middle and inner ear.To some extent it also allows us to address potentials forimpact. For marine mammals it is necessary to bring bothforms of data, direct from behavioural tests and indirect frommodels, to bear. Befare beginning those discussions, however,it is necessary to explain a few of the "rules" for sound inwater versus air.

times faster anel, at each frequency, the wavelength is 4.5 timesgreater, than in air.

How do these physical differences affect hearing? Mamma­lian ears are primarily sound intensity detectors. Intensity, likefrequency, depends on sound speed anel, in turn, on density.Sound intensity (I) is the acoustic power (P) impinging on asurface perpendicular to the direction of sound propagation,or power/unit area (I = P/a). Intensity for an instantaneoussound pressure for an outward travelling plane wave in termsof pressure, sound speed, and density is defined mathemati­callyas

L

I =pv = p..E. = E...rc rc

The combined factor (re) is the characteristic impedance ofthemedium. If we take into account the differences in soundspeed in air c = 340 m s' versus sea water c = 1530 m s'; andin density which in air = 0.0013 g cm' versus sea water = 1.03g cm'

SOUND IN AIR VERSUS WATER

Hearing is simply the detection of sound. "Sound" is the pro­pagation of a mechanical disturbance through a medium. Inelastic media like air and water, that disturbance takes theform of acoustic waves. Basic measures of sound are fre­quency, speeel, wavelength, and intensity. Frequency, meas­ured in cyc1es per second or Hertz (Hz), is defined as

F = eil

2

I = Pair 340ms-1 O.0013gcm 3 0.442mgs-lcrn-3

where c = the speed of sound (m s') and 1 is the wavelength(m).

The speed of sound is not invariable; it depends upon thedensity ofthe medium. Because water is denser than air, soundin water travels faster and with less attenuation than sound inair. Sound speed in air is approximately 340 m s'. Soundspeed in sea water averages 1530 m s' but will vary with anyfactor affecting density and any ocean region can have a highlyvariable sound profile that may change both seasonally andregionaIly. For practical purposes, in water sound speed is 4.5

To examine the sensory implications of these numbers, con­sider a hypothetical mammal that hears equally weIl in waterand in air. For this to be true, an animal would require thesame acoustic power/unit area in water as in air to have anequal sound percept, or (I,;, = Iw, ,,, )

2 2I = Pair = PWater I

arr 0.442mgs-1cm-3 (1575mgs-1crn-3 = water

81

2 23565.4Pair = Pwater

59.7 Pair =Pwater

This means the aquatic ear needs and must tolerate soundpressures in water that are ~60 times greater than are requiredin air to produce the same intensity and therefore the samesensation in the ear,

For teclmological reasons, we commonly use effective soundpressure level (SPL) rather than intensity to describe hearingthresholds (see Au 1993 for discussion). Sound pressurelevels are conventionally expressed in decibels (dB), definedas

dBSPL = IOIOg[ ~:: ]

= 2010{ ~~ )where pm is the pressure measured and pr is an arbitrary ref­erence pressure. Currently, two standardized reference pres­sures are used. For air-borne sound measures, the reference isdB re 20 jiPa rms, derived from human hearing. For und er­water sound measures, the reference pressure is dB re I jiPa.

Decibels are a logarithmic scale that depends on referencepressure. In the earlier hypothetical example, with identicalreference pressures, the animal needed a sound level 35.5 dBgreater in water than in air. However, if conventional refer­ences for measuring levels in air versus water are used, thedifferences in reference pressure must be considered as weIl.This means the underwater sound pressure level in water ifmeasured with conventional reference pressures would need tobe 61.5 dB re I pPa greater in wate r to be equivalent to thedecibel in air 01' dB re 20 pPa in air. Thus, the rule of thumb isthat to compare air versus underwater sound intensities, thenumerical value of the water sound pressure level must bethought of as being reduced by ~61.5 dB to be comparablenumerically to an intensity level reported in air.

It is important to remember that these equations describe ideal­ized comparison of air and water-borne sound. In comparingdata from different species, particularly in comparing air­based land mammal and marine mammal hearing, experi­mental condition differences are extremely important. We haveno underwater equivalent of anechoic charnbers, often resultsare obtained from one individual that may not have normalhearing, and test conditions are highly variable.

MECHANISMS OF ACOUSTIC TRAUMA

Temporary versus permanent threshold shifts

Because of our considerable interest in human hearing andhow hearing is lost or may be arneliorated, noise trauma is awell-investigated phenomenon for airborne sound. For thesake of completeness in the following discussion, noise trau­ma has been divided into lethaI and sublethaI impacts, al-

82

though only sublethaI impacts are Iikely to be relevant in thecase of long-range sonar devices. LethaI impacts are those thatresult in the immediate death or serious debilitation of themajority of animals in 01' near an intense SOUlTe, i.e., pro­found injuries related to shock wave or blast effects which arenot, technically, pure acoustic traumata. LethaI impacts arediscussed briefly at the end of this section. SublethaI impactsare those in which a hearing loss is caused by exposures tosounds that exceed the ears tolerance to some acoustic para­meter, i.e., auditory damage occurs from exhaustion or over­extension of one or more ear components. Of course, sublethaIimpacts may ultimately be as devastating as lethaI impacts,causing death through impaired foraging, predator detection,communication, stress, or mating disruption, but the potentialfor this type of extended or delayed impact from any soundsource is not weIl understood for any mamma!.

Essentially whether there is any hearing loss and, if so, whatportion of hearing is lost, comes down to three interactivefactors: intensity, frequency, and sensitivity. To determinewhether any one animal 01' species is subject to a sublethaInoise impact from a particular sound requires understandinghow its hearing abilities interact with that sound. Basically,any noise at some level has the ability to damage hearing bycausing decreased sensitivity. The loss of sensitivity is called athreshold shift. Not all noises will produce equivalent damageat some constant exposure leve!. The extent and duration of athreshold shift depends upon the synergistic effect of severalacoustic features, including how sensitive the subject is to thesound. Most recent research efforts have been directed atunderstanding the basics of how frequency, intensity, andduration of exposures interact to produce damage rather thaninterspecific differences: that is, what sounds, at what levels,for how long, 01' how often will commonly produce recover­able (TTS - Temporary Threshold Shift) versus permanent(PTS) hearing loss.

Three fundamental effects are known at this time:(1) the severity of the lass from any one signal may differamong species;(2) for pure tones, the loss centers on the incident frequency;(3) for all tones, at some balance of noise level and time, theloss is irreversible.

Hearing losses are recoverable (TTS - Temporary ThresholdShift) or permanent (PTS) primarily based on extent of innerear damage the received sound and received sound levelcauses. Temporary threshold shifts (TTS) will be broad 01'

punctate, according to source characteristics. The majority ofstudies have been conducted with cats and rodents, using rela­tively long duration stimuli (> 1 hr.) and mid to low frequen­cies (1-4 kHz) (see LEHNHARDT 1986, for summary). Inner eardamage location and severity are correlated with the powerspectrum of the signal in relation to the sensitivity of theanima!. Virtually all studies show that losses are centered nearthe peak spectra or at partial octave intervals ofthe source andare highly dependent upon the frequency sensitivity of thesubject. For narrow band, high frequency signals, losses typi­cally occur only in or near the signal band, but intensity andduration can act synergistically to broaden the loss.

The point cannot be made too strongly that this is a syner­gistic and species-specific phenomenon. Put simply, for a

sound to impact an ear, that ear must be able to hear the sound,and, equally important, the overall effect will depend on justhow sensitive that ear is to the particular sound. For this reasonthere is no single, simple number, i.e., no one sound byte, forall species that accurately represents the amount of damagethat can occur.

In effect, the duration of a threshold shift is correlated withboth the length of time and the intensity of exposure. Ingeneral, ifthe duration to intense noise is shart and the noise isnarrow, the loss is limited and recoverable. In most cases asignal intensity of 80 dB over the individual threshold at eachfrequency is required for significant threshold shifts (see Fig.I). This finding led to the current OSHA allowable limit of 90dB re 20 pPa for human warkplace exposures for broad spec­trum signals (LEHNHARDT 1986).

Unlike TTS which is highly species dependent, PTS onsets aremore general. One important aspect of PTS is that signal rise­time and duration of peak pressure are significant factars.Cornmonly, if the exposure is short, hearing is recoverable; iflong, or has a sudden, intense onset and is broadband, hearing,particularly in the higher frequencies, can be permanently lost(PTS). In humans, PTS results most often from protracted,repeat intense exposures (e.g., occupational auditory hazardsfrom background industrial noise) or sudden onset of intensesounds (e.g., rapid, repeat gun firc). Sharp rise-time signalshave been shown also to produce broad spectrum PTS at lowerintensities than slow onset signals both in air and in water(LIpSCOM 1978, LEHNHARDT 1986). Hearing loss withaging (presbycusis) is the accumulation of PTS and TTSinsults to the ear. Typically, high frequencies are lost first withthe loss gradually spreading to lower frequencies over time.

In experiments with land mammals, multi-hour exposures tonarrow band noise are used to induce both TTS and PTS andinitial shifts are often in the 10's of decibels. Work to date onmarine mammals has been much mare conservative with rela­tively short exposures that induce less than 10 dB of shiftwhich is considered invariably temporary. Consequently thereare serious concerns that the numbers from current experi­ments cannot be used to extrapolate PTS from TTS data as thecurrent curves are not yet at the conventional or comparativeTTS frontier as defined for land mammals and humans. Asnoted above, most mammals with air-adapted ears commonlyincur temporary losses when the signal is 80 dB over thresh­old. The only other available data for under-water shifts arefrom experiments that produced TTS in humans for frequen­cies between 0.7 and 5.6 kHz (our most sensitive range) fromunderwater sound sources when received levels were 150-180dB re I pPa (SMITH & WOJTOWICZ 1985, SMITH et al. 1988).Taking into account differences in measurements of soundpressure in air versus water (equations 4 and 5), these under­water levels are consistent with the 80-90 dB exposure levelsthat induce TTS in humans at similar frequencies in air.

Blast injury

Simple intensity related loss is not synonymous with blast in­jury. Acoustic trauma induced by sudden onset, loud noise (a"blast" of sound) is not synonymous with blast trauma, nor arenoise and blast effects of the same magnitude. Blast injuries

generally result from a single exposure to an explosive shockwave which has a compressive phase with a few microsecondsinitial rise time to a massive pressure increase over ambientfollowed by a rarefactive wave in which pressure drops weilbelowambient.

Blast injuries may be reparable or permanent according to theseverity of the exposure and are conventionally divided intothree groups based on severity of symptoms, which parallelthose ofbarotrauma:

MILD MODERATE SEVERErecovery partial loss permanent loss or death

Pain Otitis media Ossicular fractureand/or disloeation

Vertigo Tympanie membrane Round/ovalwindow rupture ruptureTinnitus Tympanie membrane CSF leakge

hematoma into middle earHearing Ioss Serum-blood in Cochlear and

middle ear saeeular damageTympanie tear Disseetion of mueosa

Moderate to severe stages result most often from blasts, ex­treme intensity shifts, and trauma; i.e., explosions or bluntcranial impacts that eause sudden, massive systemic pressureincreases and surges of circulatory or spinal fluid pressures(SCHUKNECHT 1993). Hearing loss in these cases results froman eruptive injury to the inner ear; i.e., with the rarefactivewave of a nearby explosion, cerebrospinal fluid pressuresincrease and the inner ear window membranes blow out due topressure increases in the inner ear fluids. Inner ear damagefrequently coincides with fractures to the bony capsule of theear or middle ear bones and with rupture of the eardrum.Although technically apressure induced injury, hearing lossand the accompanying gross structural damage to the ear fromblasts are more appropriately thought of as the result of theinability of the ear to accommodate the sudden, extreme pres­sure differentials and over-pressures from the shock wave.

At increasing distance from the blast, the effects of the shockwave lessen and even though there is no overt tissue damage,mild damage with some permanent hearing loss occurs(BURDICK 1981 in LEHNHARDT 1986). This type of loss is gener­ally called an asymptotic threshold shift (ATS) because it isthe result of saturation or in simpler terms extension past thebreaking point ofbody and certainly auditory tissues.

There is no weil defined single criterion for sublethaI ATSfrom blasts, but eardrum rupture, which is common to allstages of blast injury, has been moderately weil investigated.Although rupture per se is not synonymous with permanentloss (eardrum ruptures can repair spontaneously if less than 25% of the membrane is involved or can be repaired surgicallywith no hearing loss if greater areas are cornpromised), theincidence of tympanic membrane rupture is strongly corre­lated with distance from the blast (KERR & BYRNE 1975). Asfrequency ofrupture increases so does the incidence ofperma­nent hearing loss. In zones where >50 % tympanic membranerupture occurred, 30 % ofthe victims had long term or perma­nent loss. Trauma to other areas ofthe auditary system such asthe outer canal and middle ear bones are not nearly as weil

83

investigated. In light of concerns from the Bahamian BeakedWhale incident, this is an area warranting more research.

Concerning survivable blast trauma, in general, complex andfast-rise time sounds cause ruptures at lower overpressuresthan slow-rise time waveforms, and smaller mammals will beinjured by lower pressures larger animals. Of the animalstested to date, sheep and pig have ears anatomically closest tothose of whales and seals. The air-based data for pigs andsheep imply that overpressures 70 kPa are needed to induce100 % tympanie membrane rupture. However, cross-study/cross-species comparisons and extrapolations are risky be­cause of radically different experimental conditions as well asdifferences in acoustic energy transmission in the air andwater. The data available for submerged and aquatic animalsimply that lower pressures in water than in air induce serioustrauma (MYRICK et al. 1989; see also summary in R!CHARDSONet al. 1991). For submerged terrestrial mammals, lethal inju­ries have occurred at overpressures 255 kPa (YELVERTON 1973in MYRICK et al. 1989). In a study of Hydromex blasts in LakeErie the overpressure limit for 100 % mortality for fish was 30kPa (CHAMBERLAIN 1976). The aquatic studies imply thereforethat overpressures between 30 and 50 kPa are sufficient for ahigh incidence of severe blast injury. Minimal injury limits inboth land and fish studies coincided with overpressures of 0.5to 1 kPa.

MARINE MAMMAL HEARING

Hearing research has traditionally focused on mechanisms ofhearing loss in humans. Animal research has therefore em­phasized experimental work on ears in other species as humananalogues. Consequently we generally have investigated eithervery basic mechanisms of hearing or induced and exploredhuman auditory system diseases and hearing failures throughthese test species. Ironically, because of this emphasis,remarkably little is known about natural, habitat- and species­specific aspects of hearing in most mammals. With marinemammals we are at an extreme edge of not only habitat adap­tations but also of ear structure and hearing capabilities.

The same reasons that make marine mammals acoustically andauditorally interesting, i.e., that they are a functionally excep­tional and an aquatic adapted ear, also make them difficultresearch subjects. Marine mammal hearing has for manydecades been the poor stepchild of auditory researchprograms. Consequently, we now find ourselves for multiplereasons in need of precisely the basic research informationthat we lack. Nevertheless, we can address some issues aboutmarine mammal hearing, both directly and inferentially fromthe data in hand. While there are large gaps remaining in ourknowledge, progress has been made on some fronts related tosound and potential impacts from noise.

Marine mammals, and whales in particular, present an inter­esting hearing paradox. On one hand, marine mammal earsphysically resemble land mammal ears. Therefore, since manyforms of hearing lass are based in physical structure, it islikely hearing damage occurs by similar mechanisms in bothland and marine mammal ears. On the other hand, the sea isnot, nor was it ever, even primordially, silent. The ocean is anaturally relatively high noise environment. Principal natural

84

sound sources include seismic, volcanic, wind, and even bioticsources. Whales and dolphins in particular evolved ears thatfunction well within this context ofhigh natural ambient noise.This may mean they developed "tougher" inner ears that areless subject to hearing loss. Recent anatomical and beha­vioural studies do indeed suggest that whales and dolphinsmay be more resistant than many land mammals to temporarythresh-old shifts, but the data show also that they are subject todisease and aging processes. This means they are not immuneto hearing loss, and certainly, increasing ambient noise viahuman activities is a reasonable candidate for exacerbating oraccelerating such losses.

Unfortunately, existing data are insufficient to accuratelypredict any but the grossest acoustic impacts on marinemammals. At present, we have relatively little controlled dataon how the noise spectrum is changing in oceanic habitats as aresult of human activities. We also have little information onhow marine mammals respond physically and behaviourally tointense sounds and to long-term increases in ambient noiselevels. Our current inability to predict the impact of man-rnadesounds in the oceans has spawned serious and occasionallyvituperous debates in the scientific community as well ascostly legal battles for environmental and governmental organ­izations. Ironically, our data gaps may also be hampering thedevelopment and deployment of even simple devices such aseffective acoustic deterrents that could decrease marinemamma1by-catch.

This paper will not fill the gaps in our knowledge but ratherwill discuss our current data base on both acoustic trauma andon the sound profiles of ocean habitats in the context of whatwe know about species variations in marine mammal hearing.This section focuses on how species vary in their potential forimpact and on how we may go about determining whetherauditorially fragile species coincide with "acoustic hotspots"where man's sonic activities, particularly industrial and re­search sources in the Antarctic, must be considered to avoidactivities that could damage hearing and disrupt key beha­viours.

The data available show that all marine mammals have afundamentally mammalian ear which through adaptation tothe marine environment has developed broader hearing rangesthan are common to land mammals. Audiograms are availablefor only ten species of odontocetes (Figure 2 top) and elevenspecies (Figure 2 bottom) ofpinnipeds. All are smaller specieswhich were tested as captive animals. However, there are 119marine mammal species, and the majority are large wide­ranging animals that are not approachable or testable bynormal audiometric methods. Therefore we do not have directbehavioural or physiologie hearing data for nearly 80 % of thegenera and species of concern for coastal and open oceansound impacts. For those species for which no direct measureor audiograms are available, hearing ranges are estimated withmathematical models based on ear anatomy obtained fromstranded animals or inferred from emitted sounds and playback experiments in the wild.

The combined data from audiograms and models show thatthere is considerable variation among marine mammals inboth absolute hearing range and sensitivity. Their compositerange is from ultra to infrasonic. Odontocetes, like bats, are

-<>- Bdoga

-e- Killer Wh

";'-.tt- HAtb POlp

~--- Baiji

~100 - Bot Dolph

t: + Bot Dolph

~ ...z;:;,r Faln I<i!."- ... Rino'sD."-= ..·....Booto-= 60I/j

t:-=~

2010 100 1,000 10,000 100,000 1,000,000

F~'lueIKY (Hz)

160

C Sea llon

0 C Sea llon

- -V- N Fur Seal'!f

~ 120 +N Fur Seal~

t~

-11- Harbor

~

--- Herbor-~- Harbor= •-= 80-<I:J

j Ringed

E-4 + Harp

+ Monle

40100 1,000 10,000 100,000 1,000,000

Fre quency (Hz)

Fig. 2: Underwater audiograms for odontocctcs (top): Beluga Whale, Killer Whale, Harbour Porpoise, Bottlenosed Dolphin, False KillerWhale, Risso's Dolphin and pinnipeds, (bottom): California Sea Lion, Northem Fur Seal, Harbour Seal, Ringed Seal, Harp Seal, Monk Seal.For some species, more than one curve is shown because data reported in different studies were not consistent. Note that for both the BottlenoseDolphin and the Sea Lion, thresholds are distinctly higher for one of the two animals tested. These differences may reflect different testconditions 01' a hearing deficit in one ofthe animals, Summary data compiled from POPPER 1980, FAY 1988, Au 1993, RICHARDSON et al. 1995,

85

excellent echolocators, capable of producing, perceiving, andanalysing ultrasonic frequencies weil above any humanhearing. Odontocetes commonly have good functional hearingbetween 200 Hz and 100,000 Hz (100 kHz), although somespecies may have functional ultrasonic hearing to nearly 200kHz. The majority of odontocetes have peak sensitivities (besthearing) in the ultrasonic ranges although most have moderatesensitivity to sounds from 1 to 20 kHz. No odontocete hasbeen shown audiometrically to have acute, i.e., best sensitivityor exceptionally responsive hearing «80 dB re 1 ,LIPa) below500 Hz.

Good lower frequency hearing appears to be confined to largerspecies in both the cetaceans and pinnipeds. No mysticete hasbeen directly tested for any hearing ability, but functionalmodels indicate their functional hearing commonly extends to20 Hz, with several species, including Blue, Fin, and Bowheadwhales, that are predicted to hear at infrasonic frequencies aslow as 10-15 Hz. The upper functional range for most mysti­cetes has been predicted to extend to 20-30 kHz.

Most pinniped species have peak sensitivities between 1 kHzto 20 kHz. Some species, like the Harbour Seal, have bestsens itivities over 10kHz. Only the Elephant Seal has beenshown to have good to moderate hearing below 1kHz. Somepinniped species are considered to be effectively double-caredin that they hear moderately weIl in two domains, air andwater, but are not particularly acute in either. Others, however,are clearly best adapted for underwater hearing alone.

To summarize, marine mammals as a group have functionalhearing ranges of 10Hz to 200 kHz with best thresholds near40-50 dB re 1 ,LIPa. They can be divided into infrasonic ba­laenids (probable functional ranges of 15 Hz to 20 kHz; goodsensitivity from 20 Hz to 2 kHz; threshold minima unknown,speculated to be 60-80 dB re 1 jiPa); sonic to high frequencyspecies (100 Hz to 100 kHz; widely variable peak spectra;minimal threshold commonly 50 dB re 1 ,LIPa), and ultrasonicdominant species (200 Hz to 200 kHz general sensitivity; peakspectra 16 kHz to 120 kHz; minimal threshold commonly 40dB re l,L1Pa).

IMPACTS AND SONAR

Since the development and use of sonar in World War H, acous­tic imaging devices have been increasingly employed by themilitary, research, and commercial sectors to obtain reliable,detailed information about the oceans. On one hand, thesedevices have enormous potential for imaging and monitoringthe marine environment. On the other hand, because echo­ranging techniques involve the use of intense sound andbecause hearing is an important sensory channel for virtuallyall marine vertebrates, existing devices also represent a poten­tial source of injury to marine stocks, both predator (marinemammals) and prey. Therefore, a reasonable concern for anyeffort involving active sound use in the oceans is whether theprojection and repetition of the signals employed will adverse­Iy impact species within the "acoustic reach" of the source.Realistically, because ofthe diversity ofhearing characteristicsamong marine animals, it is virtually impossible to eliminateall acoustic impacts from any endeavour, therefore the keyissues that must be assessed are:

86

(1) What combination of frequencies and sound pressurelevels fit the task?(2) What species are present in an area the device will enso­nify at levels exceeding ambient, and(3) what are the potential impacts to those species from acous­tic exposures to the anticipated frequency-intensity combina­tions.

In order to assess potential impacts, it is necessary to obtainthe best possible estimate of the coincidence of acousticdevice parameters and auditory sensitivities for animals thatmay be exposed. Because marine mammals are both an im­portant group in terms of conservation and are generally con­sidered to be acoustically sensitive, the prirnary goal of thefollowing section is to review available data on marine mam­mal hearing and auditory systems in the context of hearingdamage mechanisms.

SUMMARY: HEARING IMPACT ISSUES

The consensus of the data is that virtually all marine mammalspecies are potentially impacted by sound sources with afrequency of 300 Hz 01' higher. Any species can be impactedby exceptionally intense sound, and particularly by intenseimpulsive sounds. However, the realistic scenario is that theeffects are a composite of three aspects: intensity, frequency,and individual sensitivity. Briefly, if you cannot hear asound01' hear it poorly, it is unlikely to have a significant effect. Ifhowever, you have acute hearing in the range of a signal, be itprop noise 01' a sonar, there is a potential for impact at agreater range than for a source you hear poorly. Because eachspecies has a unique hearing curve that differs from others inrange, sensitivity, and peak hearing, it is not possible toprovide a single number or decibel level that is safe for allspecies for all signals. It must be remembered that receivedlevels that induce acoustic trauma, at any one frequency, arehighly species dependent and are a complex interaction ofexposure time, signalonset and spectral characteristics, asweil as received versus threshold intensity for that species atthat frequency. Relatively few species are likely to receivesignificant impact for lower frequency sources. Those speciesthat currently are believed to be like1y candidates for LF acous­tic impact are most mysticetes and the Elephant Seal as theonly documented lower frequency sensitive pinniped. Mostpinnipeds have relatively good sensitivity in the 1-15 kHzrange while most odontocetes have peak sensitivities above 20kHz.

Pilot studies show that marine mammals are susceptible tohearing damage but are not necessarily as fragile as landmammals. The available data suggest that a received level ofranging typically 170-190 dB re 1 ,LIPa which is in the 80-90dB range over species-specific threshold for a narrow bandsource will induce temporary losses for hearing in and nearthat band in pinnipeds and delphinids. However, there may begreater convergence in the cross-species data depending uponhow the spectral content is assessed. The precise, critica1factors and commonality of factors across species is still to bedetermined. Estimates of levels that induce permanent thresh­old shifts in marine mammals cannot be made, at this time,except with some concern for accuracy by extrapolation fromPTS and trauma studies in land mammals.

Blasts are cardinal SOUtTes, capable of inducing broad hearinglosses in virtually all species but some resistance 01' tolerancemay occur based on body mass of marine mammals comparedto most land mammals tested.

For all devices, the question of impact devolves largely to thecoincidence of device signal characteristics with the speciesaudiogram. Because the majority of devices proposed use rela­tively low frequencies, odontocetes, with relatively poor sensi­tivity below 1 kHz as a group, may be the least likely animalsto be impacted. Mysticetes and pinnipeds have substantiallygreater potential than odontocetes for direct acoustic impactbecause of better low to mid-sonic range hearing.

Behavioural perturbations are not assessed here but a concernis noted that they are an equal or potentially more seriouselement of acoustic impacts. While auditory trauma, partic­ularly from short 01' single exposures may impair an indivi­dual, it is unlikely to impact most populations. Leng-termconstant noise that disrupts a habitat or key behaviour is morelikely to involve population level effects. In that sense, thequestion of individual hearing loss 01' animal loss forrns asingle intense exposure is far less relevant to conservation thanmore subtle, literally quieter but pervasive source that inducesbroad species loss 01' behavioural disruption.

Mitigation of any source 01' estimation of impact requires acase by case assessment and therefore suffers from the same,chronic lack of specific hearing data. To provide adequateassessments, substantially better audiometric data are re­quired from more species. To obtain these data requires aninitial three-pronged effort ofbehavioural audiograms, evokedpotentials recordings, and post-mortem examination of earsacross a broad spectrum of species. Cross-comparisons of theresults of these efforts will provide a substantially enhancedaudiometric data base and should provide sufficient data topredict all levels of impact for most marine mammals. Toachieve this goal without bias involves advocacy and fundingfrom a broader spectrum of federal and private SOUlTes. Thatin turn is likely to require a significant effort in public educa­tion about the real underlying issues that will supplant currentmisdirections 01' precipitous reactions on the part of manygroups concerned with marine conservation.

Is acoustic trauma even moderately debatable in marinemammals? Recalling the paradox mentioned earlier, there area variety of reasons to hypothesize that marine mammals mayhave evolved useful adaptations related to noise trauma. Voca­lizations levels in marine mammals are frequently cited asindicating high tolerance for intense sounds. Some whales anddolphins have been documented to produce sounds withsource levels as high as 180 to 220 dB re I pPa (RICHARDSONet al. 1991, Au 1993). Vocalizations are accepted indicators forperceptible frequencies because peak spectra of vocalizationsare near best frequency of hearing in most species, but it isimportant to recall that the two are not normally preciselycoincident.

It must be borne in mind also that animals, inc1uding humans,commonly produce sounds which would produce discomfort ifthey were received at the ear at levels equal to levels at theproduction site.

Arguments that marine mammals, simply by nature of theirsize and tissue densities, can tolerate higher intensities are notpersuasive. First, mammal ears are protected from self-gener­ated sounds not only by intervening tissues (head shadow andimpedance mismatches) but also by active mechanisms(eardrum and ossicular tensors). These mechanisms do notnecessarily provide equal protection from externally gen­erated sounds largely because the impact is not anticipated asit is in self-generated sounds.

Our active mechanisms are initiated in coordination and inanticipation of our own sound production. Just as the level of ashout is not indicative of normal 01' tolerable human hearingthresholds, source level calculations for vocalizations re­corded in the wild should not be viewed as reliable sensitivitymeasures. Further, the large head size of a whale is not acous­tically exceptional when the differences in pressure and soundspeed in water versus air are taken into account. As notedearlier, ear separation in a Bottlenosed Dolphin is acousticallyequivalent to that of a rat when the distances are corrected forthe speed of sound in water. Exactly how head size in wateraffects attenuation and even reception of incom-ing soundshas not been investigated and remains an important openquestion.

Major impacts from noise can be divided into direct physio­logic effects, such as permanent versus temporary hearingloss, and those that are largely behavioural, such as masking,aversion, or attraction. Although there is no substantial re­search accomplished in any ofthese areas in marine mammals,behavioural effects have been at least preliminarily investi­gated through playback and audiometric experiments, whilemarine mammal susceptibility to physiologic hearing loss isvirtually unexplored. Despite increasing concern over theeffects on marine mammals of man-made sound in the oceans,we still have little direct information about what soundfrequency-intensity combinations damage marine mammalears, and at present there are insufficient data to accuratelydetermine acoustic exposure guidelines for any marinemammal.

CONCLUSIONS:IMPACTS

MARINE MAMMAL ACOUSTICData from several pilot studies may, however, provide someuseful insights into both facets of the paradox. In one investi­gation (detailed below, KETTEN et al 1993, LIEN et al. 1993),ears from Humpbacks that died following underwater explo­sions had extensive mechanical trauma while animals thatwere several kilometres distant from the blasts and at the sur­face showed no significant behavioural effects. These findingsindicate adaptations that prevent barotrauma do not providespecial protection from severe auditory blast trauma, but itremains unc1ear whether lower intensity purely acousticstimuli induce temporary and/or acute threshold shifts inmarine mammals.

A second study compared inner ears from one long-term cap­tive dolphin with a documented hearing loss with the ears ofone juvenile and two young adult dolphins (KETTEN et al.1995). Studies of the oldest dolphin ears showed cellioss andlaminar demineralisation like that found in humans with pres­bycusis, the progressive sensorineural hearing loss that accom-

87

panies old age. The location and degree of neuraldegeneration implied a substantial, progressive, hearing lossbeginning in the high frequency regions, precisely the patterncommonly observed in humans. A review of the animal's be­havioural audiogram subsequently showed that over a twelveyear period this dolphin's hearing curve shifted from normalthreshold responses for all frequencies up to 165 kHz to nofunctional hearing over 60 kHz prior to his death at age 28. Forthis animal at least, the conclusion was that significant hearingloss had occurred attributable only to age-related changes inthe ear. Similar significant differences in the hearing thresh­olds consistent with age-related loss in two sea lions have alsobeen reported by KASTAK & SCHUSTERMAN (1995)

The problem of hearing loss has not been realistically con­sidered prior to this point in any systematic way in any marinemamma!. In fact, the most studied group, odontocetes, havegenerally been thought of as ideal underwater receivers. Acaptive animal 's age 01' history is not normally considered inanalysing its auditory responses, and, in the absence of overtdata (e.g., antibiotic therapy), we assume a test animal has anormal ear with representative responses for that species.

It is not clear that this is both reasonable and realistic. Particu­larly when data are obtained from one animal, it is importantto question whether that hearing curve is representative of thenormal ear for that species. The pilot studies noted aboveclearly suggest age andlor exposure to noise can significantlyalter hearing in marine mammals, and in some cases (comparethe two curves shown in Figure 2 top for BottlenosedDolphins), it is clear that some individual differences havebeen observed in "normal" captives that may be the result ofpermanent hearing loss. Further, some studies show losses inmarine mammals consistent with age-related hearing changes.Disease also complicates the assumptions that any animal hasnormal hearing, and, in parallel, too often and too facilely,there is an assumption that the only source of a marinemammal hearing loss is from exposure to anthropogenicsources.

Natural loss should be considered in any animal for whichthere is little 01' no hearing ability 01' health history. Thereforethe finding of a single animal with some hearing decrement inthe vicinity of a loud source cannot be taken as a clear indi­cator of a population level hazard from that source. On theother hand, because of the importance of hearing to theseanimals, it is also unlikely that a high incidence of loss will benormally found in any wild population, and a finding ofsubstantial hearing loss from, for instance, a mass-stranding 01'

fishery coincident with a long-term exposure to an intensesource would be appropriate cause for significant concern.

Of course, acoustic trauma is a very real and appropriatephysiologic concern. It is also one for which we can ulti­mately, given proper research, obtain data that will allow us toprovide a usable metric. That is, given that we know soundlevel X induces TTS while Y induces PTS, for frequency Z ina specific species, we can apply these data to the estimatedexposure curve for that species and determine its risk ofhearing loss. Because of the importance of hearing to marinemammals, understanding how man-made sources may impactthat sense is an important and reasonable step towards mini­mizing adverse impacts from man-made sound sources in the

88

oceans.

However, it is equally important to consider that sub-traumalevels of sound can have profound effects on individual fit­ness. These effects can take the form of masking of importantsignals, including echolocation signals, intra-species commu­nication, and predator-prey cues; of disrupting importantbehaviours through startle and repellence, 01' of acting asattractive nuisances, all of which may alter migration patterns01' result in abandonment of important habitats. Unfortunately,these issues are beyond the scope of this document as weIl asthe expertise of the author and therefore cannot be usefullydiscussed here. Nevertheless, it is important to at least note theconcern, and above all to suggest that there is a substantialneed for field monitoring of behaviours in wild populations intandem with controlled studies directed at expanding ouraudiometric data and understanding of acoustic traumamechanisms.

As indicated earlier, there are no discrete data at this time thatprovide a direct measure of acoustic impact from a calibrated,underwater sound source for any marine mamma!. Preliminarydata from work underway on captive cetaceans and pinnipeds(RIDGWAY pers. comm., SCHUSTERMAN pers. comm.) suggestthat odontocetes may have higher than typical tolerances fornoise while pinnipeds are more similar to land mammals intheir dynamic range for threshold shift effects. This responsedifference as weIl as the difference in hearing ranges - if thesedata are shown to be robust - suggest that pinnipeds are themore acoustically fragile group from most anthropogenicsound sources and that odontocetes are relatively immune 01'

require substantially higher sound levels to incur TTS.

In terms of sonars 01' in fact of any human acoustic device, theprincipal concerns are to determine a balance of frequenciesversus level versus duty cycle that will effectively detecttargets at long ranges but will not repel nor harm marinemammals within that sound field. To accomplish these goals itis necessary to determine and balance the following compo­nents:(1) What are the effective frequencies for operation, and(2) what are the hearing curves for species within the soundfield?

The fundamental concern is to avoid impact 01' harassment inthe short-term, as weIl as preventing Iong-term, multiple ex­posure effects that can compound the probability of hearingloss

For all species, the first issue in the proposed devices is signalshape, 01' rise time and peak spectra. As discussed earlier,impulsive sound has substantial potential for inducing broadspectrum, compounded acoustic trauma, i.e., an impulsivesource can produce greater threshold changes than a non-im­pulsive source with equivalent spectral characteristics. Con­sequently, impulse is a complicating feature that mayexacerbate the impact. Conventional suggestions for minimi­zing such effects are to ramp the signal, narrow the spectra,lower the pressure, and/or alter the duty cycle to allow recov­ery and decrease impact. Once again, however, it must berecalled that which if any ofthese measures is important to themarine mammal ear has not been determined. Further, it isalso important to consider the trade-offs each implies in opera-

tional effectiveness ofthe sonars in question. If decreasing oneaspect increases the parameters of another, the compositeeffects must always be kept in mind.

High intensity, ultrasonic devices of course have enormouspotential for serious impact on virtually every odontocete andtheir deployment in pelagic fisheries raises the greatestconcern after impulse or explosive sources. Such devices arerelatively unlikely, however, because they are unsuitable forlonger range detection. With high frequency sonic range de­vices, the possibility of profound impact from disruption ormasking of odontocete communication signals must certainlybe considered, as well as the possibility of coincident impactsto pinnipeds. Because the majority of devices proposed usefrequencies below ultra or high sonic ranges, odontocetes maybe the least likely to be impacted species.

Most odontocetes have relatively sharp decreases in sensitiv­ity below 2 kHz (see Fig. 3). If frequencies below 2 kHz areemployed with a non-impulsive wave-form, the potential forimpacting odontocetes is likely to be drastically reduced, but itmust also be borne in mind that it is non-zero. In every case,the difference between some to little or no significant physio­logic impact will depend upon received levels at the individualear. For the purposes of general discussion, a theoreticalcomparison is shown in Figure 3 for marine mammals audio­grams compared with a human audiogram. Because mecha­nisms and onset levels ofTTS and PTS are still unresolved formarine mammals, this curve is presented largely for the

purposes of gross comparisons of spectra of different sourceswith animal hearing ranges and is not intended to suggestmitigation guidelines. What the figure suggests is that themysticetes (which are speculated to have hearing curvessimilar to but at lower frequencies than odontocetes) and themajority of pinnipeds have substantially greater potential thanodontocetes for direct acoustic impact from low to mid-sonicrange devices. However, depending upon the diving and fora­ging patterns ofthese animals in comparison to the sound fieldpropagated by LF sonars or other devices, the risks to mysti­cetes and the majority of pinnipeds may be substantially lessthan a simple sound analysis would imply. That is, given thatsubstantialnumbers of these marine mammal groups are eithernot present or are infrequently found in the areas and depthsensonified, there is little probability of any one animalencountering a signal with an intensity and aperiod of timethat will induce acoustic trauma, despite their better absolutesensitivity to the signal.

Mitigation, like estimation of impact, requires a case by caseassessment. At this time we have insufficient data to accu­rately predetermine the underwater acoustic impact from anyanthropogenic source. Consequently, it is not possible to defi­nitively state what measures will ameliorate any one impact.

For the immediate future and in the absence of needed data, abest faith effort at mitigation must be founded on reasonedpredictions from land mammal and the minimal marinemarnmal and fish data available. It is reasonable to expect,

• Calif. Sea Uan'f N. Fur Seal

140iO-5

10-7-<i

~100GI iO-9..

I=Q",

'6'-'",- ~0

~..c:lfl

~GI....c:60~

10-13

;J.0-14

_ Harbour Seale Harbour Seal

Bottlenosed DolphinBelugawhah

Human

W False Killer "Wllale• Harbour Perpoise

10011 11111111' 1 11111111

Fr:quency (kHz) 10

1 11110.1

20 -\--..,.-...,....,.-,..,-nTr-----'1r--r-"1r-r,..,.,'TT"--r---r_rrTrTTr--r--~r_rr_n"...-_r--

0.01

Fig. 3: The human in-air audiogram and marine underwater audiograms (after FAY 1988, Au 1993, RrCHARDSON et al. 1995) were calcu1ated asw m' to allow direct camparisan with marine mammals before replotting on common SPL axes.

89

based on the similarities in ear architecture and in the shape ofbehavioural audiograms between marine and land marnmals,that marine mammals will have similar threshold shift mecha­nisms and will sustain acute trauma through similar mecha­nicalloads. Therefore, fast-rise impulse and explosive sourcesare likely to have greater or more profound impacts thannarrow band, ramped sources. Sirnilarly, we can expect that asignal that is shorter than the integration time constant of theodontocete, mysticete or pinniped ear or which has a longinterpulse interval has less potential for impact than a pro­tracted signal; however, simply pulsing the signal is not asufficient strategy without considering adequate interpulserecovery time. Strategies, such as compression, that allow thesignal to be near 01' below the noise floor are certainly worthexploring. Certainly, no single figure can be supplied for thesevalues for all species. Because of the exceptional variety inmarine mammal ears and the implications of this variety fordiversity of hearing ranges, there is no single frequency orcombination of pulse sequences that will prevcnt any impact.It is, however, reasonable, because of species-specificities, toconsider minimizing effects by avoiding overlap with thehearing characteristics of species that have the highest proba­bility of encountering the signal for each device deployed.

To that end, substantially better audiometric data are required.This means more species must be tested, with an emphasis onobtaining audiograms on younger, clearly unimpaired animalsand repeat measures from multiple animals. Too often our database has be undermined by a single measure from an animalthat may have some impairment. It is equally important toobtain some metric of the hearing impairments present innormal wild populations in order to avoid future over-esti­mates of impact from man-rnade sources. To obtain these datarequires a three-pronged effort of behavioural audiograrns,evoked potentials on live strandings, and post-mortem exarni­nation of ears to determination of thc level of "natural" diseaseand to hone predictive models ofhearing capacities.

The most pressing research need in terms of marine mammalsis data from live animals on sound parameters that inducetemporary threshold shift and aversive responses. Indirectbenefits of behavioural experiments with live captive animalsthat address TTS will also test the hypotheses that cellularstructure in the inner ear of odontocetes may be related toincreased resistance to auditory trauma. Combined data fromthese two areas could assist in determining whether or to whatextent back-projections from land mammal data are valid.

Biomedical techniques, such as ABR and functional MRI,offer considerable potential for rapidly obtaining mysticeteand pinniped hearing curves. Evoked potential studies ofstranded mysticetes are of considerable value but must alsocarry the caveat of determining how reliable is a result from asingle animal that may be physiologically compromised. Post­mortem studies should be considered on any animal that iseuthanized after an ABR with the goal of both providing dataabout the normality of the ear and supplying feedback tomodelling studies of hearing ranges. Otoacoustic emissionexperiments are not considered to be a viable approach forcetaceans; they may provide basic hearing data in pinnipedsbut are technically difficult.

Playback studies are a well-established technique but because

90

of the uncertainties about individually received levels theymay not considerably advance our knowledge of acousticimpact per se unless tied to dataloggers or very accurateassessments of the animal's sound field. Tagging and tele­metry are valuable approaches particularly if linked to field orvideo documentation of behaviour that is coordinated withrecordings of incident sound levels at the anima!. Telemetriemeasurement of physiological responses to sound, e.g., heartrate, may be valuable, but little is currently known of how tointerpret the data in terms of long-terrn impact.

Permanent threshold shift data may be obtainable by carefullydesigned experiments that expose post-rnortem marinemammal specimens to either intense sound and explosivesources since these effects are largely detectable throughphysical changes in the inner ear. These studies would alsosubstantially increase the species diversity ofthe available database because most marine mammaI species will not be testablewith conventional live animal audiometric techniques. Lastly,because many impact models depend upon assumptions aboutreceived levels at the ear, these projections would clearly beenhanced by basic measures on specimens of the underwateracoustic transmission characteristics of marine mammal headsand ears.

References and re/ated literature to hearing and hearing impactsand marine mammal hearing

Au. TVIVL. (1993): The sonar of dolphins.- Springer-Verlag, New York, N.Y.Bibikov, N.C. (1992): Auditory brainstem rcsponses in the Harbor Porpoisc

(Phocoena phoeoena).- In: l THOMAS, R. KASTELEIN & A.Y. SUPIN(eds.), Marine mammaI scnsory systems. Plenum Press, New York, 197­211.

Boenninghaus, C. (1903): Das Ohr des Zahnwales, zugleich ein Beitrag zurTheorie der Schallleitung.- Zool. Jb. 17:189-360 (not read in original) and33: 505-508.

Chamberlain, A.J (1976) The aeute and subacute effeets of underwater rockblasting, dredging, and other construetion activity on the fishes of theNantieoke region of Long Point Bay, Lake Erie.- Ontario Ministry ofNatural Resoutces. 132-143.

Clark, e IV (1990): Aeoustie behavior of mysticete whales.- In: lA.THOMAS & R.A. KASTELEIN (eds.), Sensory abilities of cetaceans:Laboratory and field evidence. Plenum Press, New York, 571-584.

Delphin, IVF (1995): Steady-state auditory-evoked potentials in three ceta­cean species elicitcd using amplitude-modulated stimuli.- In: R.A.KASTELEIN, lA. THOMAS & PE. NACHTIGALL (eds), Scnsorysystems of aguatic mammals. DeSpil Publishers, Woerden, Netherlands,25-47.

Edds, PI. (1982): Vocalizations ofthe Blue Whale, Balaenoptera musculus. inthe St Lawrence River.- Journ, Mammalogy 63: 345-347.

Edds, PI. (1988): Characteristics of Finback Balaenoptera physalus vocali­zations in the SI. Lawrenee Estuary.- Bioacoustics 1: 131-149.

Edds-Walton, PI. (1997): Acoustic communication signals of mysticetewhales.- Bioacoustics 10: 47-60.

Fay, R.R. (1988): Hearing vertebrates: A psychophysics databook.- Hill-FayAssociates Winnctka, IL.

Johnson, es. (1968). Masked tonal thresholds in the Bottlenosed POl'poise.­Journ. Acoustical Soc. Amer. 44: 965-967.

Johnson, es. (1971): Auditory masking of one pure tone by another in theBottlenosed Porpoise.- Journ, Acoustical Soc. Amer. 49: 1317-1318.

Johnson, CL (1893): Observations on the refraction and vision of the seal'seye.- Proc. Zoo I. Soc. London48: 719-723.

Kamminga, e (1988): Echolocation signal types of odontocetes.- In: PE.NACHTIGALL & PW.B. MOORE (eds.), Animal sonar processes andperformance. Plenum Press, New York, N.Y., 9-22

Kamminga, e & Wiersma, H. (1981): Iuvestigations on cetacean sonar 11.Acoustical similarities and differences in odontocete signals.- AquaticMammals 82: 41-62.

Kastak, D. & Schusterman. R.J (1995): Aerial and underwater hearing thresh­olds for 100 Hz pure tones in two pinnipcd species.- In: R.A. KASTE­LEIN, lA. THOMAS & PE. NACHTIGALL (eds.), Sensory systems of

aquatic mammals. De Spil Publishers, Woerden, Netherlands, 71-81.Kastak, D & Schusterman. R.J (1996): Temporary threshold shift in a Harbor

Seal (Phoca vitutina).- Journ. Acoustical Soc. Amer. 100(3): 1905-1908.Kaste/ein, R.A. (1991): The relationship between sensory systems and head

musculature in the Walrus.- Abstracts of the 9th Biennial Conference onthe Biology ofMarine Mammals, p. 38.

Kaste/ein, R.A. & van Gaalen, M.A. (1988): The sensitivity ofthe vibrissae ofa Paeific Walrus (Odobenus rosmarus divergens) Part 1.- AquaticMammals 14: 123-133.

Kaste/ein, R.A., Zweypfenning, R.C.VJ & Spekreijse, H. (1990): Anatomiealand histological characteristics of the eyes of a month-old and an adultHarbor Porpoise (Phocoena): In: JA. THOMAS & R.A. KASTELEIN(eds.), Sensory abilities of eetaceans: Laboratory and field evidence.Plenum Press, New York,463-480.

Ken; A.G. & Byrne, JE (1975): Concussivc cffeets ofbomb blast on thc car.­89: 131-143.

Kellogg, WN (1959): Auditory perception of submcrgcd objects byporpoises.- Journ. Acoustieal Soc. Amer, 31: 1-6.

Ketten, DR. (1992): The marine mammal ear: Specializations for aquaticaudition and echolocation.- In: D. WEBSTER, R. FAY & A. POPPER(eds.), The biology of hearing. Springer-Verlag, New York, N.Y., 717-754

Ketten, DR. (l993a): The Cetacean ear: Form, frequency, and evolution.- In: JTHOMAS, R. KASTELEIN & A. SUPIN (eds.), Marine MammalSensory Systems, Plenum Press, 53-75.

Ketten, DR. (l993b): Low frequeney tuning in marine mammaI eai-s.- Sympo­sium on Low Frequency Sound in the Ocean, Tenth Biennial Conferenceon the Biology ofMarine Mammals.

Ketten, DR. (1994): Functional analyses ofwhale ears: Adaptations für under­water hearing.- I.E.E.E Procecdings, Underwater Acoustics I: 264-270.

Ketten, DR. (1997): Structure and function in whale ears.- Bioacoustics 103­137.

Ketten, DR. & Wartzok, D (1990): Three-dimensional reconstructions of thedolphin cochlea.- In: JA THOMAS & R.A. KASTELEIN (eds.),Sensory abilities of ceraceans. Laboratory and field evidence. PlenumPress, New York, 81-105.

Ketten, DR., Lien, J & Todd, S (1993): Blast injury in Humpack Whale ears:Evidence and implications,.- Journ. Acoustical Soc. Amer. 94 (3, pt. 2):1849-1850.

Ketten, DR., Ridgway, S & Earlv, G. (1995): Apocalyptic hearing: Aging,injury, disease, and noise in marine mammal ears.- Abstracts 11th Bien­nial Conference on the Biology of Marine Mammals, p. 61.

Lehnhardt, E. (1986): Clinieal aspeets of inner ear deafness, Springer-Verlag,NewYork, NY

Liberman, M. C. (1987): Chronic ultrastructural changes in acoustic trauma:Serial-section reconstruction of stereocilia and cuticular plates.-HearingResearch 26: 65-88.

Lien, J, Todd.S; Stevick, P, Marques, F & Ketten, D (1993): The reaction ofHumpback Whales to underwater explosions: Orientation, movements,and behavior.- Journ. Acoustieal. Soe. Amer. 94 (3, pt. 2) pp. 1849.

Lipscomb, DM (1978): Noise and audiology.- University Park Press, Balti­more, MD.

Me Cormick, J G., 1J0vel; E G., Pa/in, G. & Ridgway, SH. (1970): Soundconduction in the dolphin ear.- Journ. Aeoustieal Soc. Amer. 48: 1418­1428.

McCortnick, JG., rr0vel; EG., Ridgway. SH. & Pa/in, J (1980): Sound recep­tion in the porpoise as it relates to echolocation.- In: R.G. BUSNEL & JEFISH (eds.), Animal sonar systems. Plenum Press, New York, N.Y., 449­467.

M0h/, B. (1964): Preliminary studies on hearing in seals.- Videnskab. Meddel.Dansk Naturhist. Foren. Kobenhavn 127: 283-294.

M0h/, B. (1967): Frequency discrimination in the corninon seal and a discus­sion ofthe concept ofupper hearing limit.- In: V ALBERS (ed.), Under­water acoustics. Volume Ir. Plenum Press, New York, N.Y., 43-54

Mehl, B. (1968):. Hearing in seals.- In: R. HARRISON, R. HUBBARD, R.PETERSON, C. RICE & R. SCHUSTERMAN (eds.), The behavior andphysiology of pinnipeds. Appleton Century, New York, NY, 172-195.

M0/1I, B., Surlykke, A. & Millel; L.A. (1990): High intensity Narwhal clicks.­In: JA. THOMAS & R.A KASTELEIN (eds.), Sensory abilities of ceta­ceans: Laboratory and field evidence. Plenum Press, New York, 295-303.

M{Jh/, B. & Ronald, K (1975): The peripheral auditory system of the HarpSeal, Pagophilus groenlandicus (Erxleben 1777).- Rapp. Proc.-Verbauxdes Reunions, Conseil internat. pour l' exploration de la mer 169: 516­523.

Mehl, B. & Andersen, S (1973): Echolocation: High-frequency component inthe click of thc Harbor Porpoise (Phocoena phocoena L.).- Journ. Acou­stical Soc. Amer. 57: 1368-1372.

Mehl, B., Terhune, JM. & Rona/d, K (1975): Underwater calls of the harpseal, Pagophilus groen/andicus.- Rapp. Proc.-Verbaux des Reunions,Conseil internationale pour I' exploration de la mer 169: 533-543.

Moore, PWB. (1990): Investigations on the control of echolocation pulses inthe dolphin.- In: JA. THOMAS & R.A. KASTELEIN eds.), Sensoryabilities of cetaceans: Laboratory and field evidence. Plenum Press, New

York, 305-317.Moore, PWB. &Schusterman, R.J (1976): Discrimination ofpure tonc inten­

sities by the California Sea Lion.- Journ. Acoustical Soc. Amer. 60: 1405­1407.

Moore, PWB. & Au, WWL. (1975): Underwater localization of pulsed puretones by the California Sea Lion Za/ophus californianus> Journ. Acou­stieal Soc. Amer, 58: 721-727.

Moore, PWB. & Au, WWL. (1983): Critical ratio and bandwidth of theAtlantic Bottlenose Dolphin (Tursiops truncatus).- Journ. Aeoustical Soc.Amer. Suppl. 1: 73-74.

Moore, PWB. Pawloski, Da. &Dankiewicz, L. (1995): Interaural time andintensity difference thresholds in the Bottlenose Dolphin (Tursiops trun­catus)> In: RA KASTELEIN, JA. THOMAS & PE. NACHTIGALL(eds.), Sensory systems of aquatic mummals. DeSpil Publishers,Woerden, Netherlands, 11-25.

Moore, SE & Ridgway, SH. (1995): Whistles produeed by common dolphinsfrom the southern California Bight.- Aquatic Mammals 21: 55-63.

Moore, SE., Francine, JK, Bowles, A.E. & Ford, JKB. (1988): Analysis ofcalls of Killer Whales, Orcinus orca, from Iceland and Norway.- RitFiskideildcr 11: 225-250.

Morgane, PJ & Jacobs, JS (1972): Comparative anatomy of the cetaceannervous system.- In: R.J HARRISON (cd.), Functional anatomy ofmarine mamrnals. Vol. I. Academic Press, New York, 117-224.

Mvrick, A., Cassano, E & Oliver, C. (1989): Potential for physical injury,other than hearing darnage, to dolphins from scal bombs in the YellowfinTuna purse-seine fishery: Implications from open-water tests.- NMFSreport, unpublished.

Nachtigall, PE (1986): Vision, audition, and chemoreception in dolphins andother marine mammals.- In: R..T. SCHUSTERMAN, JA. THOMAS &EG. WOOD (eds.), Dolphin cognition and behavior: A comparativeapproach. Lawrence Erlbaum Associates, Hillsdale, NJ, 79-113.

Nachtigall, PE., Au, WWL. & Pawlowski, J (1996): Low-frequency hearing inthree speeies of odontocetes.- Journ. Acoustical Soc. Amer. 100: 2611.

Nado/, JB. (1988): Quantification of human spiral ganglion cells by serialsection reconstruction and segmental density estimates.- Amer. Journ.Otolaryngology 9: 47-51.

Norris, J & Leatherwood, K (1981): Hearing in the Bowhead Whale,Ba/aena mysticetus, as estimated by cochlear morphology.- Hubbs SeaWorld Res.lnst. Tech. Rep. 81-132: 15.1-15.49.

Norris , KS, Harvev, G.W, Burzell. L.a. & Krishna Kartha, D.K (1972):Sound production in the freshwater porpoise Sotalia cf..fluviatilis Gervaisand Deville and Inia geoffrensis Blainville in the Rio Negro Brazil.- Inve­stigations on Cetacea, G. Pilleri 4: 251-262.

Norris, KS (1968): The evolution of acoustie mechanisms in odontocete ceta­ceans.- In: E.T. DRAKE (ed.), Evolution and environment. Yale Univer­sity Press, New Haven, 297-324.

Norris, KS (1969): The echolocation of marine mammals.- In: H.J AN­DERSEN (ed.): The biology of marine marnmals. Academic Press, NewYork,391-423.

Norris, KS (1980): Peripheral sound processing in odontocetes.- In: R.-G.BUSNEL & JE FISH (eds.), Animal sonar systems, Plenum Press, NewYork,495-509.

Norris, KS & Evans, WE. (1967): Directionality of echolocation clicks in theRough-Tooth Porpoise, Steno bredanensis (Lesson).-ln: W.N. TAVOLGA(ed.), Marine bio-acoustics, Vol. 2. Pergarnon Press, Oxford, 305-316.

Norris, KS & Harvey, G. W (1974): Sound transmission in the porpoisehead.- Journ. Acoustical Soc. Amer. 56: 659-664.

Norris, KS & Watkins, WA. (1971): Underwater sounds of Arctocephalusphilippii, the Juan Fernandez fur seal.- Ant. Res. Sero 18: 169-171.

Norris, KS, Würsig, B. Wells, R.S, Wiirsig, M, Brownlee, SM. Johnson, C,M.& Solow. J (1994): The Hawaiian Spinner Dolphin.- University ofCali­fornia Press, Berkeley.

Norris, KS, Prescott, JH., Asa-Doriani'V & Perkins, P (1961): An experi­mental demonstration of echolocation behavior in the porpoise, Tursiopstruncatus, Montagu.- Bio!. Bull. 120: 163-176.

Patterson, JJH. (1991): Effects of peak pressure and energy of impulses. 90:205-208.

Patton, G. W & Gerstein. E. (1992): Toward understanding mammalianhearing tractability: Preliminary acoustical perception thresholds in theWest Indian Manatee, Trichechus manatus» In: D. WEBSTER, R. FAY &A. POPPER (eds.), The biology ofhearing, Springer-Verlag, Berlin, 783.

Payne, KB., Tyack, P & Payne, R.S (1983): Progressive changes in the songsof Humpback Whales (Megaptera novaeangliae).- In: R.S. PAYNE (ed.),Communication and behavior of whales. AAAS Selected Sympos. Ser.,Westview Press, Boulder CO., 9-57.

Payne, KB., Langbauer. WJ J,: & Themas, EM (1986): Infrasonic ealls ofthe Asian Elephant (E/ephas maximusi- Behavioral Eeol. Sociobiol. 18:297-301.

Phillips, Y, Hoyt, V, Mundie, T & Dodd, K (1989): Middle ear injury inanimals exposed to complex blast waves inside an annored vehicle.- Ann.Otol. Rhinol. Laryngol. 98: 17-22.

POPOl, VV & Supin, A. Y (1990a): Electrophysiological studies on hearing in

91

some cetaceans and a manatee.- In: JA. THOMAS & R.A. KASTELEIN(eds.), Sensory abilities of cetaeeans: Laboratory and field evidenee.Plenum Press, York, 405-416.

Popov, VV & Supin, A.Y (I 990b): Loealization ofthe acoustic window at thedolphin' head.- In: JA. THOMAS & RA KASTELEIN (eds.), Sensoryabilities of cetaeeans: Laboratory and field evidenee. Plenum Press, NewYork,417-427.

Popov, VV Ladygina, TF & Supin. A.Ya. (1986): Evoked potentials of theauditory cortex of the porpoise Phocoena phocoena.- Journ. Compar.Physiol. 158: 705-711.

Popper. A.N (1980): Sound emission and detection by delphinids.- In: L.M.HERMAN (cd.), Cetacean behavior: Mechanisms and functions. JohnWiley & Sons, New York, I-52.

Renaud, DI. & Popper; A.N (1975): Sound localization by the BottlenosePorpoise Tursiops truncatus.- Journ. Exper. Biol. 63: 569-585.

Renouf, 0. (1992): Sensory reception and processing in Phoeidae and Otari­idae.- In: D. RENOUF (ed.), Behaviour ofpinnipeds. Chapman and Hall,London, 345-394.

Renouf, 0., Galway, G. & Gaborko, I. (1980): Evidence for echolocation inharbour seals.- Journ, Mal', Biol. Assoe. 60: 1039-1042.

Repenning, C (1972): Underwater hearing in seals.- In: R. HARRISON (ed.),Functional anatomy of marine mammals, Volume I. Academic Press,London, 307-331.

Reysenbach de Haan, FW (1956): Hearing in whales.- Acta OtolaryngologicaSuppl. 134: 1-114.

Richardson, WJ, Greene, c.n.», Malme, CI. & Thomson, DH (1991):Effects of noise on marine mammals.- USDIIMMA/OCS study 90-0093.LGL Eeological Research Association. Bryan, Texas.

Richardson, WJ, Greene, CR.J,:, Malme, CI. & Thomson, DH, with Moore,SE. & Würsig, B. (1995): Marine mammals and noise.- Academic Press,San Diego, CA., pp. 576.

Richmond, 0., Fleteher. E., Yelverton, J & Phillips, Y (1989): Physical corre­lates ofeardrum rupture.-Ann. Otol. Rhinol. Laryngol. 98: 35-41.

Ridgway, S & Cardei; 0. (1990): Sounds made by a neonatal Sperm Whale.­Joum. Acoustical Soc. Amer. 88, Suppl. I: 6.

Ridgway, SH (1972): Mammals of the sea: Biology and medicine.- CharlesH. Thomas, Springfield, IL.

Ridgway, SH, Bullock, TH, Cardei; DA., Seeley, R.L., Woods, 0. &Galambos, R. (1981): Auditory brainstem response in dolphins.- Proc.Nat. Acad. Sei. 78: 1943-1947.

Roberto, M. Hamemik, R. & Turrentine, G. (1989): Damage of the auditorysystem assoeiated with acute blast trauma.- Ann. Otol. Rhinol. Laryngol.Suppl. 140: 23-34.

Robineau, 0. (1969): Morphologie externe du complexe osseux temporal chezles sireniens.- Mem. Musee Nat. d'Histoire Naturelle, N.S., SeI'. A,Zoologie 60: 1-32.

Schuknecht, HF (1993): Pathology of the ear.- 2nd ed., Lea and Febiger,Philadelphia, Penn.

Shipley, C, Stewart, B.S & Bass, J (1992): Seismic communication in north­ern Elephant Seals.- In: JA. THOMAS, R.A. KASTELEIN & A.YA.SUPIN (eds.), Marine mammal sensory systems. Plenum Press, NewYork, N.Y., 553-562.

92

Smith, PF & Wojtowicz, J (1985): Temporary auditory threshold shifts in­duced by twenty-five minute continuous exposures to intense tones inwater.- Naval Med. Res. & Devel. Command, USN, Rep. 1063: 1-13.

Smith, PF, Wojtowicz, J & Carpentei; S (1988): Temporary auditory thresh­old shifts induced by ten-minute exposures to continuos tones in water.­Naval Med. Rcs. & Devel. Command, USN, Rep. 1122: 1-10.

Supin, A. Y & Popov, VV (1990): Frequency selectivity of the auditory systemof the Bottlenosed Dolphin Tursiops truncatus» In: JA. THOMAS &R.A. KASTELEIN (eds.), Sensory abilities of cetaeeans: laboratory andfield evidence. Plenum Press, NewYork, N.Y., 385-393.

Supin, A. Y & Popov, VV (1993): Direction-dependent spectral sensitivity andinteraural spectral difference in a dolphin: Evoked potential study.- Journ.Acoustical Soc. Amer. 93: 3490-3495.

Thomas, JA., Pawloski, JI. & Au, WWL. (1990): Masked hearing abilities ina false Killer Whale (Pseudorca crassidensr- In: JA. THOMAS & R.A.KASTELEIN (eds.), Sensory abilities of cetaceans: Laboratory and fieldevidence. Plenum Press, NewYork, 395-404.

Themas, JA., Zinnel, K.C & Fenn, I.M. (1983): Analysis of Wedell SealLeptonychotes weddelli vocalizations using underwater playbacks.- Cana­dian Journ, Zool. 61: 1448-1456.

Thomas, JA., Chun, N & Au, W (1988): Underwater audiogram of a falseKiller Whale (Pseudorca crassidens)» Journ. Acoustical Soc. Amer. 84:936-940.

Tyack, P (1985): An optical telemetry deviee to identify whieh dolphinproduces a sound.- Journ. Aeoustieal Soe. Amer. 85: 1892-1895.

Uexküll, J, von (1934): Astroll through the worlds of animals and men. Apicture book of invisible worlds.- In: C. SCHILLER (cd.), Instinetivebehaviour (translated 1957), London, Metheun, 5-80.

Voronov, VA. & Stosman, I.M. (1977): Frequeney-threshold eharaeteristics ofsubcortical elements of the auditory analyzer of the Phocoena phocoenaporpoise.- Zh. Evol. Biokh. I Fiziol., 6: 719. (read as English summary.

Watkins, WA. (1981): The activities and underwater sounds of Fin Whales.­Sei. Rep. Whales Res. Inst. 33: 83-117.

Watkins, WA. & Wartzok, 0. (1985): Sensory biophysies of marine mam­mals.- Marine Mammal Sei. I: 219-260.

Webster. 0., R. Fay & A. Popper (eds.) (1992): The biology of hearing.­Springer-Verlag, New York, NY.

West, CD (1985): The relationship of the spiral turns of the cochlea and thelength of the basilar membrane to the range of audible frequencies inground dwelling mammals.- Joum, Aeoustieal Soc, Amer. 77: 1091-1101.

T100d, FG. & Evans, WE. (1980): Adaptiveness and ecology of eeholoeationin toothed whales.- In: R.-G. BUSNEL & JE FISH (eds.), Animal sonarsystems, Plenum Press, New York, 381-425.

White, M.1, Norris, J, Ljungblad, 0., Baron, K. & Di Sciria, G. (1978): Audi­tory thresholds of two Beluga Whales (Delphinapterus leucas).­Hubbs/Sea World Res. Inst. and Naval Oeean Systems Center, San Diego,California, Teeh. Rep. H/SWRI 78-109.

Yelverton, J T & Richmond, 0. (1981): Underwater explosion damage riskeriteria for fish, birds, and mammals.- Journ. Aeoustieal Soe. Amer. 70(Suppl. I): p 84.

lost, WA. (1994): Fundamentals of hearing: An introduetion.- 3rd edition.Aeademie Press, New York, N.Y.